DATA AVAILABLE FOR EUGENICS RESEARCH 153 



I believe, been adequately exploited. We may assume that deviations 

 toward mental deficiency occur most frequently among families of border- 

 line or low normal intelligence. High incidence of mental deficiency in any 

 group is, therefore, an indication of much inferior intelligence in that group. 

 Army rejections for mental deficiency were intended to meet the exigencies 

 of a particular situation, but perhaps for that reason are all the more reliable 

 for our purpose. Rejections for mental deficiency by local boards at the 

 time of the war are reported by Davenport and Love. A second series of 

 rejections by camp psychiatrists are reported with racial classification by 

 Bailey and others. The latter data were never prorated. When prorated, 

 the two sets of data are complementary, because an unusually severe weed- 

 ing out by the local boards in certain states left fewer mental defects to be 

 discovered by the camp psychiatrists. The combination of these two sets 

 of data gives, I believe, the most reliable indication available of the distri- 

 bution of mental deficiency in different sections of the United S.tates. 



Another field in which detailed studies are especially needed to supplement 

 mass data is the anlysis of factors which influence sizes of families in different 

 situations. This field has been sadly neglected in eugenics research. In 

 the practice of medicine it is usually considered that prescription of reme- 

 dies should follow the diagnosis of causes. But in eugenics we have launched 

 various proposals for racial betterment without very much study of causes 

 of differential natality, which is, of course, the chief mechanism by which 

 the hereditary characteristics of a population are altered. In the study of 

 factors influencing differential natality, data on variations in birth rates in 

 relation to geographical location, time, and economic and social conditions, 

 may be very illuminating. But such mass data need to be supplemented by 

 intimate case studies of attitudes and motives and, where possible, medical 

 diagnosis of physiological capacities. 



In the use of census data and vital statistics to study reproductive trends 

 of different groups, the first consideration must be the elimination of irregu- 

 larities due to peculiar age and sex distributions. Unless such factors are 

 eliminated, results may be obtained which give an entirely false impression 

 of the vital habits of any group. For example, in the United States in 1919- 

 1920, a vital index calculated on the basis of crude birth and death rates 

 would indicate an excess of births over deaths of 89 per cent, whereas a vital 

 index calculated on a basis of age-specific death rates and maternity rates 

 per thousand women at each childbearing age gives an excess of births over 

 deaths of only 35 per cent. Very different stories are told by an excess of 

 births of 89 per cent and an excess of births of 35 per cent ! Incidentally, 

 it may be remarked that the excess of births over deaths in the United States, 



